Power cable fault type automatic identification and positioning detection method and system

Abstract

The invention discloses a power cable fault type automatic identification and positioning detection method and system, and relates to the technical field of fault detection, and the method comprises the steps: collecting an electrical parameter signal of a power cable and electromagnetic field distribution data outside a line, constructing a double-domain feature space of a cable operation state, and executing the spatial-temporal feature decomposition, forming an electrical characteristic sequence and an electromagnetic field characteristic spectrum, and respectively inputting the electrical characteristic sequence and the electromagnetic field characteristic spectrum into a fault type identification module and a fault position prediction module; the fault type identification result generated by the fault type identification module and the fault position candidate area determined by the fault position prediction module are integrated through a feature cascade network, a fusion decision result is generated, and the fault type and the fault position are determined. According to the invention, the technical problem that the traditional single-domain information detection is limited in identification capability and unstable in positioning precision in a multi-type fault scene is solved, and accurate identification and high-precision positioning of fault types are realized through parallel processing and fusion decision making of electrical parameters and electromagnetic field information.

Description

Technical Field

[0001] The present invention relates to the technical field of fault detection, and specifically to an automatic identification and location detection method and system for power cable fault types. Background Art

[0002] For a long time, power cable fault detection has mainly relied on two types of technical methods: fault identification methods based on electrical parameter analysis and fault location methods based on traveling wave theory. The electrical parameter analysis method determines the fault type by monitoring the changes in the current and voltage waveforms of the cable, specifically including transient overvoltage analysis, zero-sequence current monitoring, harmonic analysis, etc.; while the traveling wave location method determines the fault point location by capturing the propagation characteristics of the electromagnetic waves generated by the fault in the cable, mainly including the pulse reflection method, the correlation method, and the traveling wave velocity method. With the increase in the complexity of the power system, cable faults show the characteristics of diversification and concealment, and the traditional detection methods based on a single information source are facing severe challenges. Research institutions at home and abroad have proposed various improvement schemes for this problem, such as transient signal analysis based on wavelet transform, fault mode recognition based on support vector machines, multi-sensor collaborative location, etc., but these methods still have not fundamentally broken through the limitations of a single information domain, lack the effective integration of electrical parameters and electromagnetic field information, and cannot construct a complete dual-domain feature space.

[0003] The existing technology has obvious deficiencies in the field of power cable fault detection. First, the fault identification method based on a single electrical parameter has limited identification ability in multi-type fault scenarios, especially for special types such as high-resistance faults and multi-point faults, and the identification accuracy is significantly reduced. Its identification principle mainly relies on the matching degree between the fault current characteristics and the preset pattern. When the fault types present similar electrical characteristics, it is easy to cause misjudgment. Second, the accuracy of the traditional traveling wave location method is unstable due to various factors, including environmental interference, signal attenuation, wave velocity change, etc. Especially at the cable joints, due to the reflection signal aliasing caused by the wave impedance change, the location accuracy is greatly reduced. Third, the existing technology has limited ability to decompose the spatio-temporal characteristics of power signals, and it is difficult to effectively distinguish the transient component and steady-state component characteristics, as well as the zero-sequence component and positive-negative sequence component characteristics, which affects the accuracy of fault feature extraction. In addition, the existing methods generally adopt a serial processing mode, that is, first judge the fault type, and then select a suitable location algorithm according to the fault type. This processing method not only prolongs the fault diagnosis time, but also makes the location result overly dependent on the accuracy of the fault type judgment. Once the fault type is misidentified, it will lead to the wrong selection of the location algorithm, ultimately affecting the overall detection effect. More critically, the existing technology lacks an effective feature cascading network mechanism, and cannot realize the optimal integration and fusion decision of the fault type recognition result and the fault location candidate area information, and it is difficult to make full use of the complementarity of the two types of information, thus affecting the accuracy and reliability of fault detection. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed.

[0005] Therefore, the present invention provides a method and system for automatically identifying and locating power cable fault types, which can solve the problems mentioned in the background art.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: A method and system for automatically identifying and locating power cable fault types, including: collecting electrical parameter signals inside the power cable and electromagnetic field distribution data outside, and constructing a dual-domain feature space of the cable operating state; performing spatio-temporal feature decomposition on the dual-domain feature space to form an electrical feature sequence and an electromagnetic field feature map, inputting the electrical feature sequence into a fault type identification module, and inputting the electromagnetic field feature map into a fault location prediction module; the fault type identification module and the fault location prediction module constitute a dual discrimination system; the fault type identification result generated by the fault type identification module and the fault location candidate area determined by the fault location prediction module are integrated through a feature cascade network to generate a fusion decision result, and the fault type and fault location are determined; the feature cascade network realizes the optimal integration of the results of the two modules of the fault type identification module and the fault location prediction module through weighted information fusion.

[0007] As a preferred solution of the method for automatically identifying and locating power cable fault types according to the present invention, wherein: constructing the dual-domain feature space of the cable operating state includes: analyzing the temporal characteristics of the electrical parameter signals, identifying the feature change points, and dividing the electrical parameter signals into multiple feature segments; performing feature extraction on each of the multiple feature segments respectively, and connecting the features of each feature segment through a continuity constraint function to form a complete electrical parameter feature expression; synchronously obtaining the electromagnetic field distribution data corresponding to the electrical parameter signals, and constructing an electromagnetic field parameter feature expression; aligning the electrical parameter feature expression and the electromagnetic field parameter feature expression in a four-dimensional tensor structure in space and time to form a dual-domain feature space.

[0008] As a preferred solution of the method for automatically identifying and locating power cable fault types according to the present invention, the method includes: performing spatio-temporal feature decomposition on the dual domain feature space, including: performing short-time Fourier transform on the electrical parameter signals to obtain a frequency domain feature spectrum, extracting the main frequency components to form a spectrum feature set; determining a feature subspace based on the spectrum feature set, and performing non-negative tensor decomposition on the feature subspace to obtain basic electrical domain features; determining an associated region in the electromagnetic field distribution data according to the basic electrical domain features, and performing short-time Fourier transform on the associated region to obtain electromagnetic domain spectrum features; combining the basic electrical domain features with the electromagnetic domain spectrum features to form a complete spatio-temporal feature representation, and forming the electrical feature sequence and the electromagnetic field feature map.

[0009] As a preferred solution of the method for automatically identifying and locating power cable fault types according to the present invention, the implementation of the dual discrimination system includes: the fault type identification module initially identifies the fault type based on the electrical feature sequence and generates a fault feature template; the fault location prediction module receives the fault feature template, performs matching analysis with the electromagnetic field feature map, and determines the region with the highest matching degree as the fault location candidate area; the fault location prediction module feeds back the spatial distribution features of the fault location candidate area to the fault type identification module; the fault type identification module confirms or corrects the determination of the fault type through feature consistency verification according to the fed-back spatial distribution features and in combination with the phase relationship and harmonic component distribution at the corresponding position in the electrical feature sequence, and forms a fault type identification result.

[0010] As a preferred solution of the method for automatically identifying and locating power cable fault types according to the present invention, the feature cascade network includes an input layer, an intermediate layer, a decision layer, and a residual connection channel connecting the intermediate layer and the decision layer; the residual connection channel transmits the fault type identification result and the fault location candidate area to the decision layer for the decision layer to generate the fusion decision result.

[0011] As a preferred solution of the automatic identification and location detection method for power cable fault types according to the present invention, wherein: the residual connection channel transmits the fault type identification result and the fault location candidate area to the decision-making layer, including: extracting the transient component feature and the steady-state component feature in the fault type identification result, and extracting the zero-sequence component feature and the positive-negative sequence component feature in the fault location candidate area; calculating the symmetry component correlation degree between the transient component feature and the zero-sequence component feature, and comparing it with a first preset threshold; when the symmetry component correlation degree is lower than the first preset threshold, determining the module with a larger ratio of zero-sequence current to positive-sequence current in the fault type identification module and the fault location prediction module, and transmitting the corresponding result generated by this module to the decision-making layer, and at the same time screening out the feature data containing the traveling wave characteristic frequency from the other module and transmitting it to the decision-making layer; when the symmetry component correlation degree is higher than or equal to the first preset threshold, transmitting the fault type identification result and the fault location candidate area to the decision-making layer.

[0012] As a preferred solution of the automatic identification and location detection method for power cable fault types according to the present invention, wherein: after receiving the fault type identification result and the fault location candidate area, the decision-making layer performs the following processing to generate the fusion decision result: constructing a fault type feature space and a fault location feature space, and respectively mapping the fault type identification result and the fault location candidate area; for a single-phase ground fault, constructing a three-component feature space including zero-sequence component, positive-sequence component, and negative-sequence component, and calculating the similarity between the current fault feature and the standard single-phase ground fault template in the three-component feature space to determine the specific type of the fault; for an interphase short-circuit fault, constructing a double-gradient feature space including electromagnetic field gradient and electric field gradient, and calculating the similarity between the current fault feature and the standard interphase short-circuit fault template in the double-gradient feature space to determine the specific location of the fault; establishing an association mapping function between the fault type feature space and the fault location feature space, and the association mapping function describes the spatial distribution characteristics corresponding to different fault types; using the association mapping function to calculate the fusion decision result and output the fault type and the fault location.

[0013] To further solve the above technical problems, the present invention provides the following technical solution: An automatic fault type identification and location detection system for power cables, comprising: a data acquisition module, configured to collect electrical parameter signals inside the power cable and external electromagnetic field distribution data, and construct a dual-domain feature space of the cable operating state; a feature decomposition module, configured to perform spatio-temporal feature decomposition on the dual-domain feature space to form an electrical feature sequence and an electromagnetic field feature map, and input the electrical feature sequence into a fault type identification module and the electromagnetic field feature map into a fault location prediction module; a feature concatenation module, configured to integrate the fault type identification result generated by the fault type identification module and the fault location candidate area determined by the fault location prediction module through a feature concatenation network to generate a fusion decision result and determine the fault type and fault location.

[0014] A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that when the processor executes the computer program, the steps of the above-mentioned method for automatically identifying and locating faults in power cables are implemented.

[0015] A computer-readable storage medium, having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the above-mentioned method for automatically identifying and locating faults in power cables are implemented.

[0016] Advantages of the present invention: The present invention solves the technical problems that the identification ability of traditional single-domain information detection is limited in multi-type fault scenarios and the positioning accuracy is unstable. Through parallel processing and fusion decision of electrical parameters and electromagnetic field information, accurate identification of fault types and high-precision positioning are achieved, and the following technical effects are obtained: On the one hand, the dual-domain feature space provides more comprehensive fault characterization information, enabling the present invention to distinguish similar fault types that are difficult to distinguish by traditional methods; on the other hand, the integration of the feature concatenation network optimally fuses the information in the electrical domain and the electromagnetic domain, improving the anti-interference ability of fault location; in addition, in typical difficult-to-identify scenarios such as high-resistance faults, multi-point faults, and cable joint faults, the present invention effectively shortens the fault diagnosis time and reduces the outage loss of the power system. Description of the Drawings

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts.

[0018] Figure 1 It is a schematic diagram of the overall process of a method for automatically identifying and locating faults in power cables proposed by the present invention;

[0019] Figure 2 This is a diagram of a computer device in a method for automatically identifying and locating power cable fault types proposed by the present invention. Detailed implementation manners

[0020] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following will describe the detailed implementation manners of the present invention with reference to the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0022] Example 1. Refer to Figure 1 , which is an embodiment of the present invention, and provides a method for automatically identifying and locating power cable fault types.

[0023] In related technologies, the fault identification method based on a single electrical parameter has limited identification ability in multi-type fault scenarios. The identification principle mainly relies on the matching degree between the fault current characteristics and the preset pattern. When the fault types present similar electrical characteristics, it is easy to cause misjudgment. Secondly, the accuracy of traditional traveling wave positioning methods is affected by various factors and is unstable, including environmental interference, signal attenuation, wave velocity change, etc. Especially at the cable joints, due to the reflection signal aliasing caused by the wave impedance change, the positioning accuracy is greatly reduced. Thirdly, the existing technologies have limited ability to decompose the spatio-temporal characteristics of power signals, and it is difficult to effectively distinguish the transient component characteristics from the steady-state component characteristics, as well as the zero-sequence component characteristics from the positive and negative sequence component characteristics, which affects the accuracy of fault feature extraction. In addition, the existing methods generally adopt a serial processing mode, that is, first judge the fault type, and then select a suitable positioning algorithm according to the fault type. This processing mode not only prolongs the fault diagnosis time, but also makes the positioning result overly dependent on the accuracy of the fault type judgment. Once the fault type is misidentified, it will lead to the wrong selection of the positioning algorithm, ultimately affecting the overall detection effect. More critically, the existing technologies lack an effective feature cascade network mechanism, and cannot realize the optimal integration and fusion decision of the fault type identification result and the fault location candidate area information, and it is difficult to make full use of the complementarity of the two types of information, thus affecting the accuracy and reliability of fault detection.

[0024] This application provides a solution that can effectively address the above-mentioned problems. Next, multiple embodiments will be combined to elaborate in detail on how to implement the automatic identification and location detection method for power cable fault types.

[0025] Figure 1 The overall flowchart of an automatic identification and location detection method for power cable fault types is shown, including the following steps:

[0026] S1: Collect the electrical parameter signals inside the power cable and the electromagnetic field distribution data outside, and construct a dual-domain feature space for the cable operating state.

[0027] Specifically, the dual-domain feature space contains two complementary information sets, namely the electrical domain and the electromagnetic domain.

[0028] Specifically, constructing the dual-domain feature space for the cable operating state includes:

[0029] Analyze the temporal characteristics of the electrical parameter signals, identify the feature change points, and divide the electrical parameter signals into multiple feature segments;

[0030] Perform feature extraction on each of the multiple feature segments, and connect the features of each feature segment through a continuity constraint function to form a complete electrical parameter feature representation;

[0031] Synchronously obtain the electromagnetic field distribution data corresponding to the electrical parameter signals, and construct an electromagnetic field parameter feature representation;

[0032] Align the electrical parameter feature representation and the electromagnetic field parameter feature representation spatiotemporally in a four-dimensional tensor structure to form a dual-domain feature space.

[0033] The technical solution of the present invention will be described in detail below with specific embodiments.

[0034] In step S1, collect the electrical parameter signals inside the power cable and the electromagnetic field distribution data outside, and construct a dual-domain feature space for the cable operating state. The dual-domain feature space contains two complementary information sets, namely the electrical domain and the electromagnetic domain. The electrical parameter signals mainly include the voltage and current waveform data at both ends of the cable, which are obtained by high-precision sampling devices installed at both ends of the cable line; the electromagnetic field distribution data is collected by an electromagnetic field sensor array arranged along the cable route, including the spatial distribution information of the electric field strength and the magnetic field strength.

[0035] The dual-domain feature space is represented by a four-dimensional tensor structure, including a time dimension, a cable spatial position dimension, an electrical parameter dimension corresponding to the electrical domain, and an electromagnetic field parameter dimension corresponding to the electromagnetic domain, where the electrical parameter dimension corresponds to the electrical parameter signals and the electromagnetic field parameter dimension corresponds to the electromagnetic field distribution data.

[0036] Constructing the dual-domain feature space for the cable operating state specifically includes the following four steps:

[0037] First, analyze the timing characteristics of the electrical parameter signals, identify the characteristic change points, and divide the electrical parameter signals into multiple characteristic segments. When a power cable fails, its electrical parameter signals (such as current and voltage) usually exhibit obvious change characteristics. Through the designed characteristic change point detection algorithm, these change characteristic points can be automatically identified, thereby dividing the continuous electrical parameter signals into multiple characteristic segments with different characteristics.

[0038] In this embodiment, the characteristic change point detection adopts the sliding window method to calculate the change rate of the local statistical characteristics of the electrical parameter signals. When the change rate exceeds the preset threshold, this moment is marked as the characteristic change point. The specific calculation formula is as follows:

[0039]

[0040] Where, V i represents the statistical feature vector calculated within the time window i, including statistical quantities such as the mean, variance, skewness, and kurtosis of the signal; Δ i represents the normalized difference value between adjacent windows; ε is the minimum precision value of the system. When Δ i is greater than the preset threshold τ, the boundary between window i and i + 1 is marked as the characteristic change point, thereby dividing the electrical parameter signals into multiple characteristic segments.

[0041] Second, perform feature extraction on multiple characteristic segments respectively, and connect the features of each characteristic segment through the continuity constraint function to form a complete electrical parameter feature expression. For different characteristic segments, different feature extraction strategies are adopted: for the normal operation segment, steady-state features are mainly extracted; for the transient change segment, transient features are mainly extracted; for the steady-state fault segment, both the steady-state features after the fault and the transient decay features are concerned.

[0042] To ensure the continuity and smooth transition between different characteristic segments, the present invention adopts a continuity constraint function, and its expression is:

[0043]

[0044] Where, g i and g i+1 respectively represent the characteristic functions of adjacent characteristic segments; t f and t b respectively represent the end time of the previous segment and the start time of the next segment; γ and η are weight coefficients used to balance the importance of position continuity and gradient continuity, and are usually adjusted according to the specific application scenario; represents the gradient operator. By minimizing the continuity constraint function, the optimal connection method of the features of each characteristic segment can be obtained to form a complete electrical parameter feature expression.

[0045] In the third step, synchronously obtain the electromagnetic field distribution data corresponding to the electrical parameter signals, and construct the electromagnetic field parameter feature representation. When a power cable fails, the electromagnetic field distribution generated has obvious spatial characteristics, which can provide key information for fault type identification and location positioning. The electromagnetic field distribution data is obtained through a multi-point electromagnetic sensor array arranged along the cable, and each sensor measures the local electric field strength and magnetic field strength.

[0046] For the measured electromagnetic field data, extract its spatial distribution characteristics, gradient characteristics, spectral characteristics, and time-varying characteristics, and construct the electromagnetic field parameter feature representation. The electromagnetic field gradient characteristic is calculated as follows:

[0047]

[0048] where E m is the electric field strength; H m is the magnetic field strength vector; (x, y, z) are the three-dimensional space coordinates; is the gradient operator; and respectively represent the partial derivatives with respect to the three directions of x, y, and z. The electromagnetic field gradient characteristic directly reflects the spatial change rate of the electromagnetic field and is of great significance for fault location positioning.

[0049] In the fourth step, align the electrical parameter feature representation and the electromagnetic field parameter feature representation in a four-dimensional tensor structure in space-time to form a dual-domain feature space. Space-time alignment is a key step in constructing the dual-domain feature space, which ensures the consistency of the electrical parameter features and the electromagnetic field parameter features in the time and space dimensions.

[0050] When performing space-time alignment, first establish the mapping relationship between the physical position of the cable and the electromagnetic field sampling points. For the alignment of spatial positions, a piecewise linear interpolation method is used, and the expression is as follows:

[0051]

[0052] where M(l) represents the parameter value at position l on the cable, l j and l j+1 represent the positions of adjacent sampling points. Through this interpolation method, the electrical parameters and electromagnetic field parameters collected at different spatial positions can be mapped to a unified spatial coordinate system to achieve the alignment of the spatial dimension.

[0053] For the alignment of the time dimension, it is achieved through synchronous sampling clocks or post-data processing. In practical applications, the data from different sampling sources are aligned to a unified time axis using timestamp information.

[0054] Through the above four steps, the present invention constructs a dual-domain feature space for the operating state of the cable. This feature space integrates information from the electrical domain and the electromagnetic domain, providing rich feature representations for subsequent fault type identification and location positioning. Compared with traditional single-information-source methods, the dual-domain feature space can more comprehensively describe cable fault characteristics, improving the accuracy and reliability of fault detection.

[0055] S2: Perform spatio-temporal feature decomposition on the dual-domain feature space to form an electrical feature sequence and an electromagnetic field feature map. Input the electrical feature sequence into the fault type identification module and the electromagnetic field feature map into the fault location prediction module.

[0056] Among them, the fault type identification module and the fault location prediction module constitute a dual discriminant system.

[0057] Specifically, the implementation of the dual discriminant system includes:

[0058] The fault type identification module preliminarily identifies the fault type based on the electrical feature sequence and generates a fault feature template;

[0059] The fault location prediction module receives the fault feature template, performs matching analysis with the electromagnetic field feature map, and determines the region with the highest matching degree as the fault location candidate area;

[0060] The fault location prediction module feeds back the spatial distribution characteristics of the fault location candidate area to the fault type identification module;

[0061] The fault type identification module, according to the fed-back spatial distribution characteristics, combines the phase relationship and harmonic component distribution at the corresponding positions in the electrical feature sequence, and verifies the feature consistency to confirm or correct the determination of the fault type, forming the fault type identification result.

[0062] The spatio-temporal feature decomposition adopts a method combining tensor decomposition and Fourier transform.

[0063] Specifically, performing spatio-temporal feature decomposition on the dual-domain feature space includes:

[0064] Perform short-time Fourier transform on the electrical parameter signal to obtain the frequency-domain feature spectrum, and extract the main frequency components to form a spectrum feature set;

[0065] Determine the feature subspace based on the spectrum feature set, and perform non-negative tensor decomposition in the feature subspace to obtain the basic electrical domain features;

[0066] Determine the associated region in the electromagnetic field distribution data according to the basic electrical domain features, and perform short-time Fourier transform on the associated region to obtain the electromagnetic domain spectrum features;

[0067] Combine the basic electrical domain features with the electromagnetic domain spectrum features to form a complete spatio-temporal feature representation, and form an electrical feature sequence and an electromagnetic field feature map.

[0068] The technical solution of the present invention will be described in detail below in conjunction with specific embodiments.

[0069] Spatio-temporal feature decomposition adopts a method combining tensor decomposition and Fourier transform. Tensor decomposition is used to process the multi-dimensional data structure of the dual-domain feature space, while Fourier transform is used to extract the frequency-domain features of the signal. Through this combined method, the characteristic manifestations of power cable faults in the time and space dimensions can be effectively captured.

[0070] Specifically, spatio-temporal feature decomposition of the dual-domain feature space includes the following four steps:

[0071] In the first step, perform short-time Fourier transform on the electrical parameter signal to obtain the frequency-domain feature spectrum, and extract the main frequency components to form the spectrum feature set. The short-time Fourier transform can reflect the frequency characteristics of the signal at different time periods and is particularly suitable for analyzing the transient process during cable faults.

[0072] The mathematical expression for performing short-time Fourier transform on the electrical parameter signal x(t) is:

[0073]

[0074] where X(t, ω) represents the result of the short-time Fourier transform of the signal x(t) at time t and angular frequency ω, w(τ - t) is a window function centered at t, and e -jωτ is the Fourier kernel function. By calculating |X(t, ω)| 2 the time-frequency spectrum of the signal is obtained, and the frequency component with the most concentrated energy is extracted as the main frequency component to form the spectrum feature set F s ={f 1 , f 2 ,..., f K}, where f k represents the k-th main frequency component.

[0075] In the second step, determine the feature subspace based on the spectrum feature set, and perform non-negative tensor decomposition in the feature subspace to obtain the basic electrical domain features. First, determine the feature subspace according to the spectrum feature set, and then perform non-negative tensor decomposition in this subspace to extract the basic features of the electrical domain.

[0076] Let the dual-domain feature space be represented as a four-dimensional tensor where I 1 , I 2 , I 3 and I 4They respectively represent the dimensions of the time dimension, the spatial position dimension, the electrical parameter dimension, and the electromagnetic field parameter dimension. Based on the spectral feature set, the eigen-subspace S in the tensor T can be determined. Perform non-negative tensor decomposition on the eigen-subspace:

[0077]

[0078] where represents the outer product of vectors, represents the basis vector of the r-th component in the n-th dimension, and R is the decomposition rank. Through this decomposition, the basic feature representation of the electrical domain is obtained

[0079] In the third step, according to the basic features of the electrical domain, the associated regions in the electromagnetic field distribution data are determined, and the short-time Fourier transform is performed on the associated regions to obtain the spectral features of the electromagnetic domain. Using the correlation between the basic features of the electrical domain and the electromagnetic field distribution data, the regions in the electromagnetic field data that are most relevant to the fault are determined, and frequency domain analysis is performed on these regions.

[0080] The correlation between the basic features of the electrical domain and the electromagnetic field data can be quantified by calculating the correlation coefficient:

[0081]

[0082] where p represents the spatial position on the cable, M(p) represents the electromagnetic field data at position p, Cov represents the covariance, and σ M(p) represent the standard deviations of the basic features of the electrical domain and the electromagnetic field data respectively. The regions where the correlation coefficient ρ(p) is greater than the threshold θ are determined as the associated regions Ω = {p|ρ(p) > θ}.

[0083] Perform the short-time Fourier transform on the electromagnetic field data of the associated regions to obtain the spectral features M of the electromagnetic domain f .

[0084] In the fourth step, combine the basic features of the electrical domain and the spectral features of the electromagnetic domain to form a complete spatio-temporal feature representation, forming an electrical feature sequence and an electromagnetic field feature map. Through the feature fusion method, the features of the electrical domain and the electromagnetic domain are combined to form a complete spatio-temporal feature representation.

[0085] Electrical feature sequence is formed by arranging the basic features of the electrical domain in chronological order, reflecting the characteristic changes of the cable fault in the time dimension:

[0086]

[0087] where represents the basic features of the electrical domain at time t i .

[0088] Electromagnetic field characteristic spectrum M g It is formed by the distribution of electromagnetic domain spectrum characteristics in the spatial dimension, reflecting the characteristic distribution of faults in spatial positions:

[0089] M g ={M f (p 1 ), M f (p 2 ),..., M f (p L )};

[0090] Among them, M f (p j ) represents the electromagnetic domain spectrum characteristics at position p j .

[0091] The fault type recognition module and the fault location prediction module constitute a dual discrimination system. This dual discrimination system improves the accuracy of fault recognition and location through the mutual cooperation and feedback between the two modules.

[0092] Specifically, the implementation of the dual discrimination system includes the following four steps:

[0093] First step, the fault type recognition module initially recognizes the fault type based on the electrical feature sequence and generates a fault feature template. By analyzing the feature patterns in the electrical feature sequence, the possible fault types are initially judged, and the corresponding feature templates are generated for each possible fault type.

[0094] The fault feature template is an abstract representation of the typical performance of a specific type of fault in the electrical feature space, containing the key feature indicators of this type of fault and their change patterns. For n possible fault types, the generated set of fault feature templates is T = {T 1 , T 2 ,..., T n}, where T i represents the feature template of the i-th fault type.

[0095] Second step, the fault location prediction module receives the fault feature template, performs a matching analysis with the electromagnetic field characteristic spectrum, and determines the region with the highest matching degree as the fault location candidate area. The fault location prediction module projects the fault feature template into the electromagnetic field feature space, calculates the matching degree with the electromagnetic field characteristic spectrum, and determines the most likely fault location area.

[0096] For the fault feature template T i , its matching degree calculation formula with the electromagnetic field characteristics at position p is:

[0097]

[0098] Among them, M g (p) represents the electromagnetic field characteristics at position p, · represents the vector inner product, and ||·|| represents the vector norm. The area with the highest matching degree S(T i , p) is determined as the fault position candidate area A i ={p|S(T i , p)>λ i}, where λ i is the matching threshold.

[0099] In the third step, the fault position prediction module feeds back the spatial distribution characteristics of the fault position candidate area to the fault type recognition module. The spatial distribution characteristics of the fault position candidate area include information such as the position range of the candidate area, the electromagnetic field intensity distribution, and the gradient change. These information can provide important basis for the further confirmation of the fault type.

[0100] The spatial distribution characteristics can be represented by the vector D i ={d 1 , d 2 ,..., d m}, where d j represents the j-th spatial feature parameter of the fault position candidate area. These feature parameters are fed back to the fault type recognition module for subsequent fault type confirmation.

[0101] In the fourth step, the fault type recognition module, based on the fed-back spatial distribution characteristics, combines the phase relationship and harmonic component distribution at the corresponding position in the electrical feature sequence, and through feature consistency verification, confirms or corrects the fault type determination to form the fault type recognition result. By comparing the spatial distribution characteristics with the phase relationship and harmonic distribution in the electrical feature sequence, it is verified whether the preliminarily identified fault type is consistent with all observed data.

[0102] The feature consistency verification is based on the following criteria: First, extract the phase relationship and harmonic component distribution in the electrical feature sequence corresponding to the time period of the fault position candidate area, then compare them with the expected features in the fault feature template, and calculate the consistency score. If the consistency score is higher than the preset threshold, the preliminarily identified fault type is confirmed; otherwise, the fault type determination is corrected according to the consistency analysis result.

[0103] Through the four steps of the above dual discriminant system, the present invention realizes the mutual verification and optimization of fault type recognition and fault position location, and improves the overall recognition and location accuracy. Compared with the traditional single-direction fault recognition method, the dual discriminant system can make more effective use of multi-source information and reduce the misjudgment rate.

[0104] S3: The fault type recognition results generated by the fault type recognition module and the fault location candidate areas determined by the fault location prediction module are integrated through a feature cascade network to generate a fusion decision result, and the fault type and fault location are determined.

[0105] Among them, the feature cascade network realizes the optimal integration of the results of the fault type recognition module and the fault location prediction module through weighted information fusion.

[0106] Specifically, the feature cascade network includes an input layer, an intermediate layer, a decision layer, and a residual connection channel connecting the intermediate layer and the decision layer; the residual connection channel transmits the fault type recognition results and the fault location candidate areas to the decision layer for the decision layer to generate a fusion decision result.

[0107] Specifically, the residual connection channel transmits the fault type recognition results and the fault location candidate areas to the decision layer, including:

[0108] Extract the transient component features and steady-state component features in the fault type recognition results, and extract the zero-sequence component features and positive-negative sequence component features in the fault location candidate areas;

[0109] Calculate the symmetrical component correlation degree between the transient component features and the zero-sequence component features, and compare it with the first preset threshold;

[0110] When the symmetrical component correlation degree is lower than the first preset threshold, determine the module with a larger ratio of zero-sequence current to positive-sequence current in the fault type recognition module and the fault location prediction module, and transmit the corresponding result generated by this module to the decision layer. At the same time, screen out the feature data containing the traveling wave characteristic frequency from the other module and transmit it to the decision layer;

[0111] When the symmetrical component correlation degree is higher than or equal to the first preset threshold, transmit the fault type recognition results and the fault location candidate areas to the decision layer.

[0112] Specifically, after receiving the fault type recognition results and the fault location candidate areas, the decision layer performs the following processing to generate a fusion decision result:

[0113] Construct a fault type feature space and a fault location feature space, and map the fault type recognition results and the fault location candidate areas respectively;

[0114] For single-phase grounding faults, construct a three-component feature space including zero-sequence components, positive-sequence components, and negative-sequence components, calculate the similarity between the current fault features and the standard single-phase grounding fault template in the three-component feature space, and determine the specific fault type;

[0115] For phase-to-phase short-circuit faults, a double-gradient feature space including the electromagnetic field gradient and the electric field gradient is constructed. The similarity between the current fault features and the standard phase-to-phase short-circuit fault template is calculated within the double-gradient feature space to determine the specific fault location.

[0116] An association mapping function between the fault type feature space and the fault location feature space is established. The association mapping function describes the spatial distribution characteristics corresponding to different fault types.

[0117] The association mapping function is used to calculate the fusion decision result and output the fault type and the fault location.

[0118] The technical solution of the present invention will be described in detail below in conjunction with specific embodiments.

[0119] The feature cascade network realizes the optimal integration of the results of the fault type recognition module and the fault location prediction module through weighted information fusion. The feature cascade network can not only integrate the results of the two modules, but also dynamically adjust the weights of the results of each module according to different fault situations, thereby improving the accuracy of the final decision.

[0120] Specifically, the feature cascade network includes an input layer, an intermediate layer, a decision layer, and a residual connection channel connecting the intermediate layer and the decision layer; the residual connection channel transmits the fault type recognition result and the fault location candidate area to the decision layer for the decision layer to generate a fusion decision result.

[0121] The input layer is responsible for receiving the preliminary results of the fault type recognition module and the fault location prediction module and converting them into feature vectors that can be processed by the network. The intermediate layer contains multiple processing units for extracting high-level features and performing preliminary fusion operations. The decision layer generates the final fusion decision result based on all available information.

[0122] The residual connection channel is a key component in the feature cascade network. It ensures that the original fault type recognition result and the fault location candidate area information can be directly transmitted to the decision layer, avoiding information loss that may occur during the intermediate layer processing. This design draws on the residual network structure in deep learning and effectively improves the performance and stability of the network.

[0123] Specifically, the residual connection channel transmits the fault type recognition result and the fault location candidate area to the decision layer, including the following steps:

[0124] The first step is to extract the transient component features and steady-state component features in the fault type recognition result, and extract the zero-sequence component features and positive-negative sequence component features in the fault location candidate area.

[0125] The transient component characteristics mainly reflect the instantaneous changes at the initial stage of a fault, including the initial fault waveform, rise time, oscillation characteristics, etc.; the steady-state component characteristics reflect the characteristics of the system when it reaches a relatively stable state after a fault, including the steady-state fault current, voltage amplitude, phase relationship, etc.

[0126] The zero-sequence component characteristics refer to the zero-sequence components of three-phase current or voltage, which are mainly used to judge whether there is a ground fault; the positive- and negative-sequence component characteristics reflect the symmetry of the three-phase system and are used to judge the type of interphase fault.

[0127] In the second step, calculate the correlation degree of the symmetric components between the transient component characteristics and the zero-sequence component characteristics, and compare it with the first preset threshold.

[0128] The correlation degree of symmetric components is an index to measure the correlation between the transient component characteristics and the zero-sequence component characteristics. It can be used to judge whether there is consistency between the fault type recognition result and the fault location candidate area. When the correlation degree of symmetric components is high, it indicates that the results of the two modules have good consistency; when the correlation degree of symmetric components is low, it indicates that there may be inconsistent results and further analysis and processing are required.

[0129] In the third step, when the correlation degree of symmetric components is lower than the first preset threshold, determine the module with a larger ratio of zero-sequence current to positive-sequence current in the fault type recognition module and the fault location prediction module, and transfer the corresponding result generated by this module to the decision-making layer. At the same time, screen out the characteristic data containing the traveling wave characteristic frequency from the other module and transfer it to the decision-making layer.

[0130] The ratio of zero-sequence current to positive-sequence current is an important index to judge the severity of a ground fault. When the correlation degree of symmetric components is low, selecting the result of the module with a larger ratio of zero-sequence current to positive-sequence current as the main reference can more accurately reflect the fault situation. At the same time, screening out the characteristic data containing the traveling wave characteristic frequency from the other module can provide supplementary information for fault location.

[0131] The traveling wave characteristic frequency refers to the characteristic frequency shown when the electromagnetic wave generated by a fault propagates in a cable. It is related to the distance from the fault point to the measurement point and is an important basis for fault location.

[0132] In the fourth step, when the correlation degree of symmetric components is higher than or equal to the first preset threshold, transfer the fault type recognition result and the fault location candidate area to the decision-making layer.

[0133] When the correlation degree of symmetric components is high, it indicates that the results of the fault type recognition module and the fault location prediction module have good consistency. The results of the two modules can be transferred to the decision-making layer simultaneously for the decision-making layer to comprehensively consider and generate the final fusion decision result.

[0134] Specifically, after receiving the fault type recognition result and the fault location candidate area, the decision-making layer performs the following processing to generate a fusion decision result:

[0135] First, construct a fault type feature space and a fault location feature space, and map the fault type recognition result and the fault location candidate area respectively.

[0136] The fault type feature space is a multi-dimensional feature space, and each dimension represents a characteristic parameter of a fault type, such as zero-sequence current amplitude, phase relationship, harmonic content, etc. Mapping the fault type recognition result into this feature space can more intuitively represent the relationships and differences between different fault types.

[0137] The fault location feature space is also a multi-dimensional feature space, and each dimension represents a characteristic parameter related to the fault location, such as electromagnetic field strength, gradient distribution, traveling wave time delay, etc. Mapping the fault location candidate area into this feature space can more comprehensively analyze the characteristic distribution of the fault location.

[0138] Second, for single-phase grounding faults, construct a three-component feature space containing zero-sequence components, positive-sequence components, and negative-sequence components, calculate the similarity between the current fault characteristics and the standard single-phase grounding fault template within the three-component feature space, and determine the specific fault type.

[0139] The three-component feature space is a feature space specifically designed for single-phase grounding faults, which comprehensively considers the characteristics of three symmetrical components: zero-sequence components, positive-sequence components, and negative-sequence components. In this feature space, different types of single-phase grounding faults (such as metallic grounding, arcing grounding, intermittent grounding, etc.) exhibit different characteristic distributions.

[0140] By calculating the similarity between the current fault characteristics and the standard single-phase grounding fault template, the specific type of single-phase grounding fault can be determined. The standard fault template is a typical fault characteristic representation established based on a large amount of historical fault data, which provides a reference standard for fault type determination.

[0141] Third, for interphase short-circuit faults, construct a double-gradient feature space containing electromagnetic field gradient and electric field gradient, calculate the similarity between the current fault characteristics and the standard interphase short-circuit fault template within the double-gradient feature space, and determine the specific fault location.

[0142] The double-gradient feature space is a feature space specifically designed for interphase short-circuit faults, which focuses on two characteristics: electromagnetic field gradient and electric field gradient. Interphase short-circuit faults will cause obvious changes in electromagnetic fields and electric field gradients near the fault point, and these gradient characteristics are of great significance for the accurate determination of the fault location.

[0143] By calculating the similarity between the current fault characteristics and the standard phase-to-phase short-circuit fault template, the specific location of the fault can be determined more accurately. For different types of phase-to-phase short-circuit faults (such as two-phase short circuit, three-phase short circuit, etc.), different standard fault templates can be used for matching and comparison.

[0144] In the fourth step, establish the correlation mapping function between the fault type feature space and the fault location feature space. The correlation mapping function describes the spatial distribution characteristics corresponding to different fault types.

[0145] The correlation mapping function is the bridge connecting the fault type feature space and the fault location feature space, and it describes the characteristic performance of different fault types in terms of spatial distribution. For example, single-phase grounding faults usually show concentrated zero-sequence current and small changes in the electromagnetic field gradient; while phase-to-phase short-circuit faults show obvious positive and negative sequence currents and drastic changes in the electromagnetic field gradient.

[0146] By establishing this correlation mapping relationship, the internal connection between the fault type and the fault location can be understood more comprehensively, providing theoretical support for the final fusion decision.

[0147] In the fifth step, use the correlation mapping function to calculate the fusion decision result and output the fault type and the fault location.

[0148] Based on the correlation mapping function, the decision-making layer comprehensively considers the information in the fault type feature space and the fault location feature space and calculates the final fusion decision result. This result includes two aspects: one is the determined fault type, such as single-phase grounding, two-phase short circuit, etc.; the other is the determined fault location, usually represented by the distance from the starting point of the cable.

[0149] Through these five processing steps of the feature cascade network, the present invention realizes the optimal integration of the fault type recognition result and the fault location candidate area, generates an accurate fusion decision result, and provides a scientific basis for the rapid handling of power cable faults. Compared with traditional fault diagnosis methods, the method of the present invention significantly improves the accuracy and reliability of fault recognition and location through multi-source information fusion and dual-module collaborative optimization.

[0150] In summary, the present invention solves the technical problems that the traditional single-domain information detection has limited recognition ability and unstable positioning accuracy in multi-type fault scenarios. Through the parallel processing and fusion decision of electrical parameters and electromagnetic field information, the accurate recognition and high-precision positioning of fault types are realized, and the following technical effects are achieved: on the one hand, the dual-domain feature space provides more comprehensive fault characterization information, enabling the present invention to distinguish similar fault types that are difficult to distinguish by traditional methods; on the other hand, the integration of the feature cascade network optimizes and fuses the information in the electrical domain and the electromagnetic domain, improving the anti-interference ability of fault positioning; in addition, in typical difficult-to-identify scenarios such as high-resistance faults, multi-point faults, and cable joint faults, the present invention effectively shortens the fault diagnosis time and reduces the outage loss of the power system.

[0151] Embodiment 2 is an embodiment of the present invention, which provides an automatic recognition and positioning detection system for power cable fault types, including:

[0152] Embodiment 3, referring to Figure 2 , is an embodiment of the present invention. The difference from the previous embodiment is that if the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical disks and other various media that can store program codes.

[0153] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device.

[0154] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, then editing, interpreting, or otherwise processing it as appropriate, and then storing it in a computer memory.

[0155] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), and the like.

[0156] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for automatically identifying and locating power cable fault types, characterized in that: include: Collect the electrical parameter signals inside the power cable and the external electromagnetic field distribution data to construct a dual-domain feature space of the cable operation status; Performing spatiotemporal feature decomposition on the dual-domain feature space to form an electrical feature sequence and an electromagnetic field feature spectrum, inputting the electrical feature sequence into a fault type recognition module, and inputting the electromagnetic field feature spectrum into a fault location prediction module; the fault type recognition module and the fault location prediction module constitute a dual discrimination system; The fault type identification result generated by the fault type identification module and the fault location candidate area determined by the fault location prediction module are integrated through a feature cascade network to generate a fusion decision result and determine the fault type and fault location; the feature cascade network realizes the optimized integration of the dual module results of the fault type identification module and the fault location prediction module through weighted information fusion.

2. The method for automatically identifying and locating power cable fault types according to claim 1, characterized in that: The dual-domain feature space for constructing the cable operation status includes: Analyze the timing characteristics of the electrical parameter signal, identify the characteristic change points, and divide the electrical parameter signal into multiple characteristic segments; Performing feature extraction on the plurality of feature segments respectively, and connecting the features of the feature segments through a continuity constraint function to form a complete electrical parameter feature expression; synchronously acquiring electromagnetic field distribution data corresponding to the electrical parameter signal, and constructing an electromagnetic field parameter characteristic expression; The electrical parameter characteristic expression and the electromagnetic field parameter characteristic expression are aligned in time and space in a four-dimensional tensor structure to form a dual-domain characteristic space.

3. The method for automatically identifying and locating power cable fault types according to claim 2, characterized in that: Performing spatiotemporal feature decomposition on the dual domain feature space includes: Performing short-time Fourier transform on the electrical parameter signal to obtain a frequency domain characteristic spectrum, and extracting a main frequency component to form a frequency spectrum feature set; Determine a characteristic subspace based on the frequency spectrum feature set, and perform non-negative tensor decomposition in the characteristic subspace to obtain basic features in the electrical domain; Determine the associated region in the electromagnetic field distribution data according to the electrical domain basic features, perform short-time Fourier transform on the associated region, and obtain electromagnetic domain spectrum features; The electrical domain basic features and the electromagnetic domain spectrum features are combined to form a complete spatiotemporal feature representation, thereby forming the electrical feature sequence and the electromagnetic field feature spectrum.

4. The method for automatically identifying and locating power cable fault types according to claim 3, characterized in that: The implementation of the dual discrimination system includes: The fault type identification module preliminarily identifies the fault type based on the electrical feature sequence and generates a fault feature template; The fault location prediction module receives the fault feature template, performs matching analysis with the electromagnetic field feature map, and determines the area with the highest matching degree as the fault location candidate area; The fault location prediction module feeds back the spatial distribution characteristics of the fault location candidate area to the fault type identification module; The fault type identification module confirms or corrects the fault type determination through feature consistency verification based on the spatial distribution characteristics of the feedback, combined with the phase relationship and harmonic component distribution of the corresponding position in the electrical feature sequence, to form a fault type identification result.

5. The method for automatically identifying and locating power cable fault types according to claim 4, characterized in that: The feature cascade network includes an input layer, an intermediate layer, a decision layer, and a residual connection channel connecting the intermediate layer and the decision layer; the residual connection channel transmits the fault type identification result and the fault location candidate area to the decision layer, so that the decision layer generates the fusion decision result.

6. The method for automatically identifying and locating power cable fault types according to claim 5, characterized in that: The residual connection channel transmits the fault type identification result and the fault location candidate area to the decision layer, including: Extracting transient component features and steady-state component features in the fault type identification result, and extracting zero-sequence component features and positive-negative sequence component features in the fault location candidate area; Calculating the symmetric component correlation between the transient component feature and the zero-sequence component feature, and comparing the correlation with a first preset threshold; When the correlation of the symmetrical components is lower than the first preset threshold, a module having a larger ratio of zero-sequence current to positive-sequence current in the fault type identification module and the fault location prediction module is determined, and a corresponding result generated by the module is transmitted to the decision layer, and at the same time, feature data containing a characteristic frequency of a traveling wave is screened out from another module and transmitted to the decision layer; When the correlation of the symmetric components is higher than or equal to the first preset threshold, the fault type identification result and the fault location candidate area are transmitted to the decision layer.

7. The method for automatically identifying and locating power cable fault types according to claim 6, characterized in that: After receiving the fault type identification result and the fault location candidate area, the decision layer performs the following processing to generate the fusion decision result: Constructing a fault type feature space and a fault location feature space, respectively mapping the fault type identification result and the fault location candidate area; For a single-phase grounding fault, a three-component feature space including a zero-sequence component, a positive-sequence component, and a negative-sequence component is constructed, and the similarity between the current fault feature and the standard single-phase grounding fault template is calculated in the three-component feature space to determine the specific type of the fault; For phase-to-phase short circuit faults, a dual gradient feature space including electromagnetic field gradient and electric field gradient is constructed, and the similarity between the current fault feature and the standard phase-to-phase short circuit fault template is calculated in the dual gradient feature space to determine the specific location of the fault; Establishing an association mapping function between a fault type feature space and a fault location feature space, wherein the association mapping function describes spatial distribution characteristics corresponding to different fault types; The association mapping function is used to calculate the fusion decision result, and the fault type and fault location are output.

8. A system for automatically identifying and locating power cable fault types, based on the method for automatically identifying and locating power cable fault types according to any one of claims 1 to 7, characterized in that: include, The data acquisition module is used to collect the electrical parameter signals inside the power cable and the external electromagnetic field distribution data to construct a dual-domain feature space of the cable operation status; A feature decomposition module, used to perform spatiotemporal feature decomposition on the dual-domain feature space to form an electrical feature sequence and an electromagnetic field feature spectrum, and input the electrical feature sequence into a fault type identification module and the electromagnetic field feature spectrum into a fault location prediction module; The feature cascade module is used to integrate the fault type identification result generated by the fault type identification module with the fault location candidate area determined by the fault location prediction module through a feature cascade network to generate a fusion decision result and determine the fault type and fault location.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for automatically identifying and locating power cable fault types according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for automatically identifying and locating power cable fault types according to any one of claims 1 to 7 are implemented.

Images